Invasive fungal infections (IFI) are major causes of death in high-risk haematological patients receiving induction therapy for acute leukaemia or intensified immunosuppression due to acute or chronic graft-vs.-host disease (GvHD) following allogeneic stem cell transplantation (SCT). Recently, two randomised studies showed the efficacy of a posaconazole prophylaxis (PP) in these patients to prevent IFI. This prompted the strong recommendation for the use of PP in national and international guidelines. As we started PP in our leukaemia and transplantation unit in summer 2007, we retrospectively analysed the impact of PP on the incidence of possible, probable or proven IFI in this group of patients. Incidence of IFI according to the revised EORTC criteria, published in 2008, was reviewed retrospectively in a group of high-risk patients treated in our unit 1 year before the start of PP compared with the same group in the following year with PP. First analysis was performed on an intention-to-treat basis comparing patients during 1 year of PP with the same group of patients in the year before the start of PP. In a second, deeper analysis, patients were grouped for fluconazole or posaconazole irrespective of the time period the prophylaxis was given. In a first intent-to-treat analysis, 56 patients were analysed in the period without PP (noPP) compared with 34 patients in the period with PP. Overall IFI rates (possible, probable and proven IFI) were reduced from 47% (noPP group) to 35% (PP group). In a second analysis, only patients receiving either fluconazole or PP were analysed, resulting in 29 patients in the noPP group and 36 patients in the PP group. There was a reduction in overall IFI in the PP group especially in the acute myeloid leukaemia (AML) induction patients, but this does not reach statistical significance because of low patient numbers. However, initiation of antifungal therapy was significantly less frequent in AML induction patients in the PP group compared with the noPP group. Unfortunately, this does not result in reduced mortality rates, as mortality in the PP group is higher (15% vs. 7%) than in noPP patients because of double the number of patients with severe GvHD in the PP group. Both breakthrough infections were documented in this subgroup of patients. Our data, collected in every day clinical practice, add further evidence to the advantage of a PP strategy in this group of high-risk patients. However, more data are urgently needed to assess the impact of PP on the incidence and pattern of fungal infections and the strategies to be used in patients with persisting fever and pulmonary infiltrates receiving PP, especially in the setting of SCT and GvHD.